1. Field of the Invention
The invention relates to a method of monitoring the continuous flows of data between the output of data transmission systems and the input of data processing equipment.
2. Prior Art
Such flows of data occur, for example in connection with the transmission of radar data for air traffic control systems. Therefore, the invention particularly concerns a method for providing air traffic control systems or the like with data, whose control centers are supplied with radar data via a network, whereby feeding of the radar installations into the network as well as uncoupling of the radar data for the air traffic control systems takes place via network junctions, and whereby the radar data are transmitted further via networks to integrated operating systems and, if need be, to a backup system. Furthermore, the invention relates to a device for implementing the method.
The European air traffic control organizations of the Netherlands, Belgium, Luxembourg and Germany, and the EUROCONTROL Agency jointly operate a network for supplying control centers with radar data (RADNET=Radar Data Network). Feeding of the radar installations into the RADNET system, as well as uncoupling of the radar data for the air traffic control systems from the RADNET system takes place via network junctions (RMCDE=Radar Message Conversion and Distribution Equipment). Further transmission of the radar data takes place via networks to the integrated operating systems, as well as to a backup system, which, however, only offers limited functionality. In cases of emergency, the backup system is additionally supplied with radar data via a direct tie to the radar stations.
Since the time of introduction of the RADNET system, a regional point is supplied with radar data basically in two ways: while the local radar station in the form of a all-round radar installation at the site of the airport (ASR=Airport Surveillance Radar) is directly fed to the display system via the Radar Message Conversion and Distribution Equipment (RMCDE), all other radar stations of a radar network are made available via the Radar Data Network (RADNET). Versus the method commonly employed prior to the introduction of the RADNET, according to which method all required radar stations had to be fed in directly on the respective processing system, this already constitutes a considerable simplification and thus also a saving of costs.
With the current method for medium and long range radar stations, radar data are fed into two operating facilities of the German Air Control System (DFS=Deutsche Flugsicherung), said facilities being independent of each other. For the local radar stations (ASRxe2x80x94Airport Surveillance Radar), the feed takes place in one location (xe2x80x9clocal feedxe2x80x9d). For a regional point, this means that in addition to the local ASR-station, further line stations are fed in quasi xe2x80x9clocallyxe2x80x9d if such a regional point is a RADNET feed point.
The network junction RMCDE consists of four channels, which are independent of each other, whereby each channel is capable of processing all required radar data. Two of said channels are combined in each case in one unit. Reversing between the two units is possible only manually, whereas within one unit, reversing between the channels takes place automatically in case of error. In the past, problems occurred in isolated cases with the radar supply in spite of the high redundancy of the radar message conversion and distribution equipment (RMCDE). Furthermore, total failure of the radar supply poses a basic problem, which cannot be bridged in any satisfactory manner by the xe2x80x9cemergency supplyxe2x80x9d of the existing backup system, and this not only because only radar data and no flight planning data can be transmitted with the backup system.
Therefore, it has already been proposed earlier to change the RMCDE-concept to the extent that the network junctions are separated into two independent RMCDE""s, namely an RMCDE N (=net) and an RMCDE D (=direct). The connection to the RADNET takes place via the RMCDE N, whereas all directly connected radar stations (RADNET feed points) are connected via the RMCDE D. Both RMCDE""s transmit their data simultaneously to two networks with a local area of propagation (LAN=Local Area Network), in particular the Ethernet and the FDDI networks. Thus all radar data are available either directly, or via the networks simultaneously and independently of each other to the online system and the backup system. Parallel therewith, all directly connected radar stations are still connected to the backup system via so-called ADR""s (all-purpose data replicators). In the event of total failure of all RMCDE""s (RMCDE D and RKCDE N), an xe2x80x9cemergencyxe2x80x9d feed takes place via the backup system. However, in this process, only the radar stations intended for the RADNET feed are available for the evaluations. Coverage of the entire air space of a regional point is not accomplished in this way. This would be possible only if in addition to the radar stations required for the RADNET-feed, additional radar stations were directly connected for single coverage of the air space. This, however, would ensue enormous line costs. In addition, the connection capacity of the existing radar stations would have to be expanded, which would particularly result in interface multiplication.
Furthermore, consideration has already been given to install the ADR""s for safety reasons not in the same area where the two RMCDE""s N and D are already installed, so that in the event both RMCDE""s should fail in a fire, at least an emergency feed of the system could be assured via the ADR""s by means of the directly connected radar stations. The problem of inadequate radar coverage connected with such an approach could be overcome only by leasing additional transmission lines, which would be connected with considerable costs and connection problems for the radar stations.
In order to assure at least single radar coverage both in the event of total failure of the network junction (RMCDE D and N) and partial failure of the network junction, direct connection of a corresponding number of radar stations is therefore required in connection with the present system.
The backup system employs a PC-based radar data processing system (xe2x80x9cTracViewxe2x80x9d), which, however, permits no connection to the flight plan processing system. The connection of radar stations to the backup system via the ADR""s, furthermore, leads to high investment costs and leases for the lines required if more radar stations than those required for the RADNET-feed are fed in. Single radar coverage of the air space could be safely secured only if additional, more remotely located radar stations were directly connected.
Now, the problem of the present invention is to make available a method of the type specified above, by which in the event of failure of the transmission system, the complete flow of data of the data processing equipment connected downstream is made available again within seconds via an alternative transmission path.
Said problem is solved according to the invention in that the transmitted data are counted at the output of the data transmission device in time intervals with adjustable parameters;
that a data flow is reproduced based on the counted data;
that the transmitted data are stored over several intervals and compressed into an average value;
that when a new value is received, the oldest value is in each case erased (FIFO-buffer);
that the average value of the measured data formed by means of the FIFO-buffer is used for a sensitivity curve, which generates a time window for the periodic comparison of the added data for an error detector;
that the error detector signalizes a total failure of the data to be transmitted in the predetermined time unit, and initiates the selection mode for accessing a predefined data source in a suitable data network; and
that the data received in said way are fed into the respective data supply system via interfaces.
A method is already known from EP 0 584 918 A, in connection with which the transmission of serial bit sequences is monitored via a modem. This, however, only serves the purpose of monitoring the lines. In this connection, the data (bits) are located on the image transmission layer (physical layer/OSI layer 1) and cannot be decoded. The bit data of modem channels are different in this connection from information data or useful data that are transported via the switching layer (network layer/OSI layer 1) and the transporting layer (transport layer/OSI layer 4). Only said information data, or useful data supply the flows of data between the data transmission devices and the data processing equipment, said flows of data being of interest in the present case. The bit data dealt with in EP 0 584 918 A are data for synchronization, or data for idling fields, and not information or useful data.
Also the method described in EP 0 606 682 A employs known routines for detecting the carrier frequency and the transmission error rate on the bit transmission layer and the safety layer (physical layer/OSI layer 1; and the link layer/OSI layer 2). As with the method known from EP 0 584 918 A, only the physical interruption of lines can be detected with such routines. As opposed to the above, according to the method as defined by the invention, an error can be detected also even if the line per se is flawless. Therefore, the information or useful data are checked, but not the bit data of the physical layers OSI 1 and 2.
So that failures can be detected in the radar data supply for air traffic control systems, and a substitute supply with radar data is automatically secured, provision is made according to the invention
that on the network of an air traffic control system, the radar target messages are counted in time intervals with adjustable parameters (e.g. seconds);
that a data flow is reproduced based on the radar target messages; that the measured values are stored over several intervals, and compressed into a=n average value;
that when a new value is received, the oldest value is in each case erased (FIFO-buffer);
that the average value of the radar target messages formed by means of the FIFO-buffer is used as input value for a sensitivity curve generating a time window for periodically comparing the added radar target messages for an error detector;
that the error detector signalizes a total failure of the radar target messages in the predetermined time unit, and initiates the selection mode for accessing a predetermined radar data source in the public data network; and
that the radar data received in said way are fed into the respective air traffic control system or the like via interfaces.
Even in the event of complete failure of the input of radar data, via the network junctions RMCDE of the RADNET, the feed is made possible within a time interval permitting continuous control of the air traffic, and an undisturbed operating condition is restored without doing without inputting radar stations. A significant advantage of the method as defined by the invention consists in that the radar data operationally required in the event of trouble are made available via automatically selected connections only when there is a demand for such data, instead of holding them available 24 hours per day for a contingency which possibly may never occur. While a time span of one to two antenna rotations (about 10 to 20 seconds) is usually required for safely recognizing the failure of a radar station, it is possible with the method as defined by the invention to safely detect the failure of all radar stations supplying data over the network already in the second raster. In this process, the method does not require any special infrastructure of air traffic control systems, but rather operates independently of the employed hardware.
So that the radar reproduction can be restored within seconds in the event of system errors, provision is made according to the invention that the measured values are stored over several intervals with adjustable parameters of, for example 1 s. It has been found in connection with air traffic control systems that it is advantageous if the measured values are stored over five intervals.
In order to be able to build up complete radar coverage within seconds, the radar stations are selected and the radar data are transmitted via the ISDN or another data network, in particular via a satellite-based data network. In this way, transmission costs are incurred only in a case of actual utilization. Furthermore, the system-inherent redundancies of the ISDN-network guarantee a transmission path whose safety against failure cannot be achieved with a directly connected line.
According to a further development of the idea of the invention, provision is made that the data of the radar stations selected via the ISDN or another suitable data network are received in a monitoring computer (PC), and fed into both the networks, in particular the Ethernet and the FDDI network, and into the backup system with complete circumvention of the network junction (RMCDE).
In order to assure that the radar stations required for the radar coverage are selected in due time, and that the radar data are made available, provision is made according to a further development of the invention that the monitoring logic of the monitoring computer continuously checks the flow of radar data, and that the monitoring logic of the monitoring computer, in the event the flow of data (number of target messages per time unit=self-learning threshold value) is interrupted, interprets this as a criterion for a failed network junction (RMCDE ), and starts building up the selected connections via the ISDN or another data network for all missing radar information.
As stated above, the air traffic control systems are supplied with radar data via network junctions (RMCDE). In a system where several RMCDE""s operating independently of each other form a network junction, i.e. where the network junctions may consist of one or a plurality of junction computers (RMCDE""s) operating independently of one another, for example an RMCDE N (net) for the connection to the RADNET, and an RMCDE (D (direct), via which all directly connected radar stations (RADNET feed-in points), provision is made according to the invention that for the duration of the failure of an RMCDE N and/or RMCDE D, the radar reproduction systems downstream are supplied with substitute radar data via ISDN- or other selected connections of external network junctions RMCDE""s) or radar installations. In this way, all connected radar stations are available again to the radar data processing system within seconds via selected connections (ISDN or the like), with no failure to be found in the radar reproduction. Online and backup systems are continuously supplied with radar data.
In the event of failure of the network junctions RMCDE D, provision is made according to a further development of the invention that the directly connected radar stations or an external network junction (RMCDE) are selected for the duration of the failure of the RMCDE, and that the radar data are made available by the monitoring computer to the networks, in particular to the Ethernet and FDDI networks, via interfaces. The outputs of the RMCDE D are monitored in this connection as with the RMCDE N. The functional sequences for building up the connections are identical and only differ in that other radar stations or other RMCDE""s are selected. However, in this connection, it is no longer possible to transmit the radar data to the defective RMCDE. The radar data are rather fed into the networks via interfaces.
The method as defined by the invention can be employed also for monitoring the data flow in a defined time interval. In this case, the system recognizes the difference between a data failure and a time span not monitored.
The device provided for solving the problem of the present invention for implementing the method consists of a network for supplying control centers with radar data (RADNET), into which network the radar data of radio stations are fed via network junctions (RMCDE""s), and to which network the radar data for the air traffic control systems are coupled. In this process, the radar data are further transmitted to the integrated operating systems, and, if need be, to a backup system via networks (Ethernet, FDDI-network).
Air traffic control systems have already been designed redundant since a long time, whereby the network junctions, for which substitution is to be connected, are redundant as well. For maintaining the operating condition it is eminently important how fast an error can be detected and eliminated. In connection with an air traffic control system known from the printed document: Cristian F. et al., xe2x80x9cFault Tolerance in Air Traffic Control Systemsxe2x80x9d, ACM Transactions on Computer Systems, Volume 14, No. 3, August 1996, pp 265-286, the redundancies shown there provide no clue with respect to the reversing criteria that are critical in terms of time, and as to the technical design of the monitoring modules, which would show how the redundancies have to be tied into the system, so that the overall system will operate without operational failure.
Therefore, in order to restore in the event of a failure the radar data supply an undisturbed operating condition without having to do without feeding radar stations, and without the emergency operation of the backup system in a time span ensuing from the strict requirements with respect to the reaction behavior of the system in terms of time, provision is made according to the invention that the device consists of an error detection system (=the ARTE systemxe2x88x92Automatic Radardata Transmission Equipment), which as a monitoring computer that acquires the radar data downstream of the network junctions (RMCDE""s) on the system-internal networks of the air traffic control system; that the monitoring computer compares in periods the number of incoming target messages; that the error detection system uses an average value formed by means of an FIFO-buffer as input value for a sensitivity curve generating a time window for the periodic comparison of the added radar data for an error detector; and that in the event of an error, and in particular in the event of a total failure of the radar target messages, the monitoring computer selects a remote station on the network (RADNET), or one or several radar stations via the ISDN or some other data network, and switches to the air traffic control system additionally required radar data, in particular radar target messages. With such a device it is possible to distinctly increase the quality of the radar data supply, while the costs are reduced at the same time. Furthermore, such a device can be easily adapted to the operational requirements, and it can be realized at favorable costs. Particularly the saving of line costs, the saving of investment and maintenance costs due to omission of the ADR, as well as further possibilities for achieving savings through the possible reduction of RMCDE""s employed, and thus savings in the form of reduced continuous operating costs, contribute to the favorable economy calculation.
In order to take into account both the number of radar target messages, which varies analogous to the air traffic density, and the varying number of radar installations feeding in, provision is made according to the invention that the error detector for determining the actual data load is designed self-learning. The loss of radar data is detected in this way more rapidly when the air traffic density is high, and an expanded time span is created for the selection mode in the public data network, in order to achieve the five seconds that have to be maintained overall for the error detection, and to make the radar data available on the network of the air traffic control system.
So as to be able to select the radar sources and to receive the radar data, and to be able to communicate with the networks, the network junction RMCDE, and the PC-based radar data processing system, the work station of the monitoring computer, according to the invention, has different interfaces (ISDN, Ethernet, FDDI, V 24). The type and number of interfaces ensue from the given local requirements.
The device as defined by the invention makes it possible, in the form of a further development of the idea of the invention, to transmit the radar data to authorized users for all sorts of different application even without occurrence of a case of failure, provided such users have access to a public data network. It is possible in this way to make the radar data supply available per selection modes to any desired users both in national and international areas.
Further features, advantages and application possibilities of the invention can be derived from the following description of exemplified embodiments of the invention and the drawing. All described and/or graphically represented features form by themselves, or in any desired combination, the object of the invention irrespectively of their combination in the claims or their dependency on preceding claims.